Methodological quality is underrated in systematic reviews and meta-analyses in health psychology

[Snow Leopard]

Methodological quality is underrated in systematic reviews and meta-analyses in health psychology

Isabel Oliveras,Josep-Maria Losilla, Jaume Vives

Objectives

In this paper, we compile and describe the main approaches proposed in the literature to include methodological quality or risk of bias into research synthesis. We also meta-review how the risk of bias of observational primary studies is being assessed and to what extent the results are incorporated in the conclusions of research synthesis.
Study Design and Setting

Electronic databases were searched for systematic reviews or meta-analyses related to health and clinical psychology. A random sample of 90 reviews published between January 2010 and May 2016 was examined.
Results

A total of 46 reviews (51%) performed a formal assessment of the risk of bias of primary studies. Only 17 reviews (19%) linked the outcomes of quality assessment with the results of the review.
Conclusion

According to previous literature, our results corroborate the lack of guidance to incorporate the risk of bias assessment in the results of systematic reviews and meta-analyses. Our recommendation is to appraise methodological quality according to domains of risk of bias to rate the degree of credibility of the results of a research synthesis, as well as subgroup analysis or meta-regression as analytical methods to incorporate the quality assessment.

http://www.jclinepi.com/article/S0895-4356(17)30471-7/pdf